JP2013092636A - Image processing device, microscope system, image processing method, and image processing program - Google Patents

Abstract

PROBLEM TO BE SOLVED: To appropriately correct a shading caused by an inclination of a slide glass and a shade of a foreign material in a light path.SOLUTION: An image processing device includes: a storage part 130 for storing pluralities of sample images picked up in an image field of a zone where a sample exists and of blank images picked up in an image field of a zone where no sample exists; an image separation part 151 for separating a pixel value of a pixel constituting each of plural blank images into a bias component corresponding to the inclination of the pixel value and a noise component corresponding to a noise, for at least the plural blank images including a glass image picked up in an image field of a zone where the slide glass exists, of the blank images stored in the storage part; a corrected blank image generation part 152 for generating a corrected blank image having, as components of a pixel value, a bias component that the glass image possesses and a component calculated from a noise component that each of the plural blank images possesses; and an image calibration part 153 for calibrating the sample image using the corrected blank image.

Description

  The present invention relates to an image processing apparatus, a microscope system, an image processing method, and an image processing program that perform image processing on a specimen image obtained by imaging a specimen observed with a microscope.

  In an optical device such as a camera, a phenomenon in which the amount of light in the peripheral region is usually lower than the center in a plane perpendicular to the optical axis due to the characteristics of an imaging optical system such as a lens. This phenomenon is generally called shading. For this reason, conventionally, image degradation has been suppressed by performing image processing based on empirically obtained correction values, actual measurement values, and the like on captured images. Such image processing is called shading correction.

  Similarly, shading occurs in a microscope. In particular, in the case of a microscope, it is preferable to optimize image processing for each objective lens because the characteristics of the light amount reduction in the peripheral region differ depending on the objective lens. Therefore, for example, in Patent Document 1, shading correction data is prepared for each objective lens, and image processing is performed by switching the shading correction data according to the objective lens to be used.

  However, according to such a method, since a shading correction pattern must be prepared for each objective lens in advance, the processing is slightly complicated. In addition, the current shading state may not correspond to a previously prepared shading correction pattern due to environmental changes such as a light source misalignment. Therefore, in Patent Document 2, a shading correction pattern is created based on an image signal acquired by capturing an image of an imaging region in a state where the sample does not exist in the imaging region, and the imaging region is obtained in a state where the sample exists in the imaging region. An image signal obtained by capturing the image is corrected with the shading correction pattern. Accordingly, it is not necessary to prepare a shading correction pattern for each objective lens in advance, and shading correction according to environmental changes can be applied to the image.

JP 2001-292369 A JP 2004-272077 A

  By the way, in the microscope, when the slide glass is placed on the stage, if the optical axis and the slice glass are not orthogonal but slightly inclined, shading may occur in the inclination direction. In such a case, it is possible to correct the shading in the tilt direction by setting the imaging area to an area where the sample does not exist on the slide glass, and using the shading correction pattern based on the acquired image signal. it can.

  In a microscope, when a foreign substance such as dust exists in the optical path, the shadow of the foreign substance may appear in the image. In such a case, by using the shading correction pattern created by the method disclosed in Patent Document 2, the shadow of the foreign matter can be erased by correction.

  However, if a foreign object is present on the slide glass and the shading correction pattern is created by setting the imaging area to an area on the slide glass where the sample does not exist, the foreign object present in the imaging area at this time is corrected for shading. Reflected in the pattern, it affects the shading correction on the image of the sample. In order to avoid this, if the shading correction pattern is created by setting the imaging area to an area where the slide glass itself does not exist, it becomes impossible to correct the shading caused by the inclination of the slide glass.

  The present invention has been made in view of the above, and is an image processing apparatus, a microscope system, an image processing method, and an image processing apparatus that can appropriately correct shading caused by the inclination of a slide glass and the shadow of a foreign object in an optical path. An object is to provide an image processing program.

  In order to solve the above-described problems and achieve the object, an image processing apparatus according to the present invention captures a sample image captured in an imaging field of a region where a sample exists and an imaging field of view of an area where the sample does not exist. A plurality of blank images, and a plurality of blank images including a glass image captured in an imaging field of view of an area where a slide glass exists among the blank images stored in the storage unit. On the other hand, the glass image has an image separation unit that separates pixel values of pixels constituting each of the plurality of blank images into a bias component corresponding to the inclination of the pixel value and a noise component corresponding to noise. A correction bracket that creates a correction blank image having a bias component and a component calculated from a noise component of each of the plurality of blank images as a component of a pixel value And click the image creation unit, characterized in that it comprises an image correction unit for calibrating the sample image using the correction blank image.

  In the image processing apparatus, the plurality of blank images include, in addition to the glass image, a white image captured in an imaging field of view in a region where the slide glass does not exist, and the correction blank image creation unit includes the glass image The correction blank image is created from the bias component included in and the component calculated from the noise component included in the glass image and the white image.

  In the image processing apparatus, the plurality of blank images include a plurality of glass images, and the correction blank image creation unit extracts a noise component common to each other from the noise components of the plurality of glass images. The correction blank image is created from the common noise component and at least one of the bias components respectively included in the plurality of glass images.

  In the image processing apparatus, when the noise component extraction unit has a difference in noise component value in a predetermined pixel corresponding to a predetermined threshold value between two glass images of the plurality of glass images, One of the noise component values in a predetermined pixel is set as the common noise component value.

  In the above image processing apparatus, the corrected blank image creation unit sets an average value of noise components in pixels corresponding to the plurality of glass images as the value of the common noise component.

  In the image processing apparatus, the image separation unit separates pixel values of pixels constituting each of the plurality of blank images into the bias component and the noise component by using a median filter process. .

  In the image processing apparatus, the image separation unit separates pixel values of pixels constituting each of the plurality of blank images into the bias component and the noise component using rolling ball processing. .

  In the image processing apparatus, the correction blank image creation unit configures the correction blank image from an approximate value that approximates the bias component included in the glass image and a noise component included in each of the plurality of blank images. The pixel value of the pixel is calculated.

  The image processing apparatus further includes a correction blank image size changing unit that changes the size of the correction blank image by performing an interpolation process based on the magnification of the microscope, and the image calibration unit changes the correction blank image size. The specimen image is calibrated using the corrected blank image whose size has been changed by the unit.

  In the image processing apparatus, at least one of the image separation unit, the correction blank image creation unit, and the image calibration unit executes processing for each band of the specimen image.

  The microscope system according to the present invention images the imaging field through the optical processing system, a stage on which the slide glass can be placed, an optical system provided to face the stage, and the optical system. An image acquisition unit that acquires an image, and a field change unit that changes the imaging field of view by moving at least one of the stage and the optical system in a direction orthogonal to the optical axis of the optical system. It has the microscope which has.

  The microscope system causes the microscope to perform imaging by aligning the imaging field of view with a region where the sample does not exist and acquire a blank image, and to perform imaging while aligning the imaging field of view with a region where the sample exists. It further has a control part which performs control which acquires a sample picture.

  In the microscope system, the control unit causes the microscope to image the sample image and the blank image with the same depth of focus.

  In the microscope system, the control unit moves the sample on the microscope apparatus while relatively moving the stage on which the sample is placed and the optical system in a direction perpendicular to the optical axis of the optical system. The plurality of sample images are captured by imaging, the image calibration unit calibrates each of the plurality of sample images, and the image processing apparatus connects the calibrated sample images to each other. It further has a virtual slide generator for generating a virtual slide image.

  In the microscope system, the microscope device further includes a magnification changing unit that changes a magnification of an image to be captured.

  The image processing method according to the present invention includes a storage step of storing a plurality of sample images captured in an imaging field of view of a region where a sample is present and blank images captured in an imaging field of view of the region where the sample is not present, Of the blank images stored in the storage step, at least for a plurality of blank images including a glass image captured in an imaging field of view where a slide glass exists, pixels of each of the plurality of blank images An image separation step for separating a pixel value into a bias component corresponding to the gradient of the pixel value and a noise component corresponding to noise, a bias component included in the glass image, and noise included in each of the plurality of blank images Correction blank image creation step for creating a correction blank image having a component calculated from the components as a component of the pixel value , Characterized in that it comprises an image calibrating step of calibrating the sample image using the correction blank image.

  An image processing program according to the present invention stores a plurality of sample images captured in an imaging field of view of an area where a sample exists and blank images captured in an imaging field of view of an area where the sample does not exist, Of the blank images stored in the storage step, at least for a plurality of blank images including a glass image captured in an imaging field of view where a slide glass exists, pixels of each of the plurality of blank images An image separation step for separating a pixel value into a bias component corresponding to the gradient of the pixel value and a noise component corresponding to noise, a bias component included in the glass image, and noise included in each of the plurality of blank images A corrected blank image creation process for creating a corrected blank image having a component calculated from the components as a pixel value component And-up, characterized in that to execute the image calibration step of calibrating the sample image to the computer using the correction blank image.

  According to the present invention, pixel values of pixels constituting each of a plurality of blank images including at least a glass image are separated into a bias component and a noise component, and the bias component included in the glass image and each of the plurality of blank images A correction blank image is created using the noise component as a pixel component, and the sample image is calibrated using the correction blank image, so that shading caused by the tilt of the slide glass and the shadow of foreign matter in the optical path are corrected appropriately. It becomes possible to do.

FIG. 1 is a schematic diagram showing a configuration example of a microscope system according to Embodiment 1 of the present invention. FIG. 2 is a schematic diagram schematically showing the configuration of the microscope apparatus shown in FIG. FIG. 3A is a schematic view of the stage shown in FIG. 2 as viewed from the objective lens side. FIG. 3B is a schematic view of the stage shown in FIG. 2 as viewed from the objective lens side. FIG. 3C is a schematic view of the stage shown in FIG. 2 viewed from the objective lens side. FIG. 4 is a block diagram showing the operation of the microscope system shown in FIG. FIG. 5 is a flowchart showing a process for acquiring a plurality of blank images in the first embodiment. FIG. 6 is a schematic diagram illustrating an example of a process of creating a corrected blank image from a plurality of blank images. FIG. 7 is a flowchart showing a correction blank image creation process according to the first embodiment. FIG. 8 is a schematic diagram for explaining the median filter processing. FIG. 9A is a graph showing pixel values in a one-dimensional region in a white image. FIG. 9B is a graph showing a bias component in a one-dimensional region in the white image. FIG. 9C is a graph showing noise components in a one-dimensional region in the white image. FIG. 10A is a graph showing pixel values in a one-dimensional region in the glass image. FIG. 10B is a graph showing a bias component in a one-dimensional region in the glass image. FIG. 10C is a graph showing a noise component in a one-dimensional region in the glass image. FIG. 11 is a graph showing pixel values in a one-dimensional region in the corrected blank image. FIG. 12 is a schematic diagram illustrating an example of a process of calibrating a specimen image with a corrected blank image. FIG. 13 is a schematic diagram illustrating a configuration example of a microscope system according to the second embodiment of the present invention. FIG. 14 is a flowchart showing a process for acquiring a plurality of corrected blank images in the second embodiment. FIG. 15 is a schematic diagram illustrating an example of a process of creating a corrected blank image from a plurality of glass images. FIG. 16 is a flowchart showing a correction blank image creation process according to the second embodiment. FIG. 17 is a schematic diagram showing a configuration example of a microscope system according to Embodiment 3 of the present invention. FIG. 18 is a flowchart showing a correction blank image creation process according to the third embodiment. FIG. 19 is a schematic diagram for explaining the principle of bilinear interpolation calculation. FIG. 20 is a schematic diagram illustrating a configuration example of a microscope system according to Embodiment 4 of the present invention. FIG. 21 is a flowchart showing the operation of the microscope system shown in FIG. FIG. 22 is a schematic diagram illustrating an example of a process of calibrating an enlarged specimen image with a correction blank image. FIG. 23 is a schematic diagram showing a configuration example of a microscope system according to Embodiment 5 of the present invention. FIG. 24 is a schematic diagram for explaining a method of creating a virtual slide image.

Hereinafter, embodiments according to the present invention will be described in detail with reference to the drawings. Note that the present invention is not limited to these embodiments.
(Embodiment 1)
FIG. 1 is a schematic diagram showing a configuration of a microscope system according to Embodiment 1 of the present invention. As shown in FIG. 1, a microscope system 1 according to Embodiment 1 controls a microscope apparatus 10 and an operation of the microscope apparatus 10 and an image processing apparatus 11 that processes an image acquired by the microscope apparatus 10. With.

  FIG. 2 is a schematic diagram schematically showing the configuration of the microscope apparatus 10. As shown in FIG. 2, the microscope apparatus 10 includes a substantially C-shaped arm 100, a stage 101 attached to the arm 100, and a holding unit that movably holds a slide glass on which a sample is arranged on the stage 101. 101a, an objective lens 102 provided on one end side of the lens barrel 103 so as to face the stage 101 via the trinocular tube unit 106, an image acquisition unit 104 provided on the other end side of the lens barrel 103, And a visual field changing unit 105 that changes an imaging visual field of the image acquisition unit 104. The trinocular tube unit 106 branches the observation light of the sample incident from the objective lens 102 into an image acquisition unit 104 and an eyepiece unit 107 described later. The eyepiece unit 107 is for the user to directly observe the specimen. In the following, the optical axis direction of the objective lens 102 is referred to as a Z-axis direction, and a plane orthogonal to the Z-axis direction is referred to as an XY plane. In FIG. 2, the microscope apparatus 10 is installed so that the main surface of the stage 101 substantially coincides with the XY plane.

  The objective lens 102 is attached to a revolver 108 that can hold a plurality of objective lenses (for example, the objective lens 102 ′) having different magnifications. By rotating the revolver 108 and changing the objective lenses 102 and 102 ′ facing the stage 101, the magnification of the image captured by the image acquisition unit 104 can be changed.

  Inside the lens barrel 103, a zoom unit including a plurality of zoom lenses and a drive unit (none of which is shown) for changing the positions of these zoom lenses is provided. The zoom unit enlarges or reduces the imaging target in the imaging field by adjusting the position of the zoom lens.

The image acquisition unit 104 includes, for example, an image sensor such as a CCD, and for example, captures a color image having pixel levels (pixel values) in each band of R (red), G (green), and B (blue) in each pixel. It is a possible camera. The image acquisition unit 104 receives light (observation light) emitted from the objective lens 102 and incident through the optical system in the lens barrel 103, generates image data corresponding to the observation light, and generates the image processing device 11. Output to.
Note that the image acquisition unit 104 is not limited to a three-band camera that can capture images with the three primary colors of RGB, and may be a camera that can capture images with four or more bands.

  Note that a liquid crystal tunable filter, an acoustic tunable filter, or a plurality of narrow band filters may be provided in the image acquisition unit 104 to configure a multispectral camera capable of acquiring a multispectral image.

The visual field changing unit 105 includes, for example, a motor 105a, and changes the imaging visual field by moving the slide glass held by the holding unit 101a in the XY plane.
3A to 3C are schematic views of the stage 101 as viewed from the objective lens 102 side, and the vicinity of the center of the stage 101 facing the objective lens 102 is an imaging field of view VF. As shown in FIGS. 3A to 3C, the visual field changing unit 105 includes a scanning unit 105b that is attached to the holding unit 101a and moves the holding unit 101a in the XY plane in conjunction with the motor 105a (see FIG. 2).

  The visual field changing unit 105 changes the relative position of the imaging visual field VF with respect to the slide glass SG by scanning the holding unit 101a with the scanning unit 105b. For example, FIG. 3A shows a state where the imaging visual field VF is at a position deviated from the slide glass SG. FIG. 3B shows a state in which the imaging field of view VF overlaps the area on the slide glass SG excluding the specimen SP. Furthermore, FIG. 3C shows a state in which the imaging visual field VF overlaps the sample SP.

  Instead of moving the holding unit 101a side with respect to the observation optical system, the holding unit 101a may be fixed and the observation optical system side may be moved in the XY plane to change the imaging field of view. Alternatively, both the slide glass SG and the observation optical system may be relatively moved.

  As illustrated in FIG. 1, the image processing apparatus 11 includes an input unit 110 that receives an instruction and information input to the image processing apparatus 11, and a sample image (hereinafter referred to as a sample image) acquired by the image acquisition unit 104. A display unit 120 that displays other information, a storage unit 130, a control unit 140 that controls the operation of each unit of the microscope apparatus 10 and the image processing apparatus 11, and a predetermined image for an image acquired by the microscope apparatus 10 And an arithmetic unit 150 that performs processing.

  The input unit 110 includes input devices such as a keyboard, various buttons, and various switches, and pointing devices such as a mouse and a touch panel. The input unit 110 receives signals input via these devices and inputs the signals to the control unit 140.

  The display unit 120 is realized by a display device such as an LCD (Liquid Crystal Display), an EL (Electro Luminescence) display, or a CRT (Cathode Ray Tube) display, and displays various screens according to control signals output from the control unit 140. indicate.

  The storage unit 130 includes a semiconductor memory such as flash memory, RAM, and ROM that can be updated and recorded, a recording medium such as a hard disk, MO, CD-R, and DVD-R that is built-in or connected by a data communication terminal, and the recording medium. This is realized by a reading device or the like that reads recorded information. The storage unit 130 stores the image data output from the image acquisition unit 10, various programs executed by the control unit 140, and various setting information. Specifically, the storage unit 130 stores an image processing program 131 that calibrates a specimen image acquired by the microscope apparatus 10.

The control unit 140 and the calculation unit 150 are realized by, for example, reading various programs stored in the storage unit 130 into hardware such as a CPU.
The control unit 140 transfers instructions and data to each unit of the image processing apparatus 11 based on various data stored in the storage unit 130 and various types of information input from the input unit 110, so that the entire image processing apparatus 11 is controlled. Control the overall operation.

  In addition, the control unit 140 controls the microscope apparatus 10 to acquire a sample image and control an operation to acquire a blank image used to create a corrected blank image for calibrating the sample image. Here, the blank image is an image obtained by imaging the imaging visual field VF in accordance with an area where the specimen SP does not exist, and is a white image obtained by imaging an area where the slide glass SG does not exist (see FIG. 3A). The image and the glass image which imaged the area | region (refer FIG. 3B) on the slide glass SG except the sample SP are included.

  Specifically, the control unit 140 includes a blank image acquisition instruction unit 141 that instructs acquisition of a blank image, a sample image acquisition instruction unit 142 that instructs acquisition of a specimen image, and these blank image acquisition instruction unit 141 and the specimen image. The imaging control unit 143 causes the microscope apparatus 10 to perform imaging in accordance with an instruction from the acquisition instruction unit 142, and the visual field change control unit 144 performs control to change the imaging visual field VF in the microscope apparatus 10 according to the instruction. Among these, the blank image acquisition instruction unit 141 includes a white image acquisition instruction unit 141a that gives an instruction to acquire the above-described white image and a glass image acquisition instruction unit 141b that gives an instruction to acquire the above-described glass image.

  The arithmetic unit 150 performs predetermined image processing such as white balance processing, demosaicing, and gamma conversion on the image corresponding to the image data stored in the storage unit 130, and calibrates the sample image subjected to these processings The arithmetic processing to be performed is performed. More specifically, the calculation unit 150 includes an image separation unit 151, a corrected blank image creation unit 152, and an image calibration unit 153.

  The image separation unit 151 performs a process of separating the blank image into a bias component and a noise component for each of the plurality of blank images among the images stored in the storage unit 130. Here, the bias component is a component representing the inclination (bias) of the pixel value in the blank image, and is used for the characteristics of the observation optical system (each lens, etc.) of the microscope apparatus 10 and the inclination with respect to the optical axis L of the slide glass SG. It is a component corresponding to the variation in the amount of light caused. On the other hand, the noise component is a component caused by foreign matter (dust, dust, etc.) that has entered the observation optical system (such as between the condenser lens and the objective lens) of the microscope apparatus 10 or foreign matter attached to the slide glass SG. The image separation unit 151 includes a white image separation unit 151a and a glass image separation unit 151b, and performs the separation process on both the white image and the glass image.

The correction blank image creation unit 152 creates a correction blank image using the bias component and noise component of each blank image separated by the image separation unit 151.
The image calibration unit 153 calibrates the specimen image using the created correction blank image.

Next, the operation of the microscope system 1 will be described. FIG. 4 is a flowchart showing the operation of the microscope system 1.
First, in step S10, the microscope system 1 acquires a plurality of blank images.

FIG. 5 is a flowchart showing details of the processing for acquiring a plurality of blank images.
First, in step S101, when the white image acquisition instructing unit 141a outputs an instruction to acquire a white image, the visual field change control unit 144 responds to the imaging field of view VF with the slide glass SG. It moves to the area | region which does not (refer FIG. 3A).

In subsequent step S102, the imaging control unit 143 causes the image acquisition unit 104 to image the imaging field of view VF. Thereby, a white image is acquired. An image M _W1 shown in FIG. 6 shows an example of a white image acquired in this way. This image M _W1 is temporarily stored in the storage unit 130.

  In step S103, when the glass image acquisition instruction unit 141b outputs an instruction to acquire a glass image, the visual field change control unit 144 responds to the slide glass in which the specimen SP is not present in the imaging visual field VF. Move to a region on SG (see FIG. 3B).

In subsequent step S104, the imaging control unit 143 causes the image acquisition unit 104 to image the imaging field of view VF. Thereby, a glass image is acquired. At this time, the depth of focus may be set in accordance with the later-described imaging of a specimen image. Image M _G1 shown in FIG. 6 shows an example of the thus obtained glass image. This image M _G1 is temporarily stored in the storage unit 130.
Thereafter, the process returns to the main routine.

In step S11, the calculation unit 150 creates a corrected blank image using a plurality of blank images.
FIG. 7 is a flowchart showing details of the correction blank image creation process.
In step S111, the white image separation unit 151a reads the pixel value W of each pixel constituting the white image from the storage unit 130 for each band z (z is R, G, or B), and converts it into a bias component and a noise component. Execute the process to be separated. Hereinafter, the pixel value of the band z in the pixel located at the coordinates (x, y) of the white image is denoted as W (x, y, z) or the like. If the image acquisition unit 104 can capture images with four or more bands, the pixel value W of each band is read out.

More specifically, first, the white image separation unit 151a performs median filter processing on each pixel of the white image. Here, as shown in FIG. 8, the median filter processing refers to a pixel P _i (i = 1 to n) within a predetermined area (5 × 5 pixels in FIG. 8) around the pixel of interest P. This is a process of outputting the median value of the pixel values. Hereinafter, the output value of the median filter processing for the pixel corresponding to the pixel value I (x, y, z) is referred to as Med (I (x, y, z)) or the like.

As a result, the output value of the median filter processing for the pixel corresponding to the pixel value W (x, y, z) is the bias component W _bias (x, y, z) of the pixel value W (x, y, z). .
W _bias (x, y, z) = Med (W (x, y, z)) (1)

The noise component W _noise (x, y, z) of the pixel value W (x, y, z) is equal to the pixel value W (x, y, z) and the bias component W as shown in the following equation (2). given by the difference from _bias (x, y, z).
W _noise (x, y, z) = W (x, y, z) −W _bias (x, y, z) (2)

FIG. 9A shows a pixel value W (x, y, z) in a certain one-dimensional region in the white image. FIG. 9B shows the bias component W _bias (x, y, z) calculated for the same region. Furthermore, FIG. 9C shows the noise component W _noise (x, y, z) calculated for the same region. 9A to 9C, the horizontal axis indicates the arrangement order of the pixels arranged in a line, and the vertical axis indicates the normalized pixel value (intensity). The image M _WB1 shown in FIG. 6, the bias component W _bias of the white image extracted from the image _{M W1 (x, y, z} ) is an image corresponding to the image M _WN1 is extracted from the image M _W1 It is an image corresponding to the noise component W _noise (x, y, z). Among these, noises N1 and N2 are reflected in the image M _WN1 .

In subsequent step S112, the glass image separation unit 151b reads the pixel value G of each pixel constituting the glass image from the storage unit 130 for each band z (z is R, G, or B), and the bias component and the noise component. Execute the process of separating them. The details of this separation process are the same as those described in step S111. Thereby, the bias component G _bias (x, y, z) and the noise component G _noise (x, y, z) of the pixel value G (x, y, z) constituting the glass image are expressed by the following equation (3) and Given by (4).
G _bias (x, y, z) = Med (G (x, y, z)) (3)
G _noise (x, y, z) = G (x, y, z) −G _bias (x, y, z) (4)

FIG. 10A shows a pixel value G (x, y, z) in a one-dimensional region in the glass image. FIG. 10B shows the bias component G _bias (x, y, z) in the one-dimensional region. Further, FIG. 10C shows the noise component G _noise (x, y, z) in the one-dimensional region. 10A to 10C, the horizontal axis indicates the arrangement order of the pixels arranged in a line, and the vertical axis indicates the normalized pixel value (intensity). An image M _GB1 shown in FIG. 6 is an image corresponding to the bias component G _bias (x, y, z) of the glass image extracted from the image M _G1 , and the image M _GN1 is extracted from the image M _G1. It is an image corresponding to the noise component G _noise (x, y, z). Among these, noises N3 to N5 are reflected in the image M _GN1 .

Here, when the image M _WN1 is compared with the image M _GN1 , the noises N1 and N3 and the noises N2 and N5 appear in common although the imaging fields of view are different. _Therefore , these noises originate from the observation optical system. This is considered to be noise (shadows of foreign matter that have entered between the lenses). On the other hand, since the noise N4 appears only in the glass image, it can be considered as a shadow of the foreign matter adhering to the slide glass SG.

Further, in step S113, the corrected blank image creation unit 152 _calculates the following expression (5) from the bias component G _bias (x, y, z) of the glass image and the noise component W _noise (x, y, z) of the white image. Is used to calculate a corrected blank image G ^. The symbol G ＾ indicates that the symbol “＾ (hat)” is attached to the symbol G.

An image _MCB1 shown in FIG. 6 is a corrected blank image created in this way. FIG. 11 shows a pixel value G ^ (x, y, z) in a certain one-dimensional region in the corrected blank image. As shown in FIGS. 6 and 11, the corrected blank image M _CB1 is obtained by shading caused by the inclination of the slide glass SG with respect to the optical axis L, and noise derived from the observation optical system (for example, noise, N1, N2, N3). , N5), only the influence exerted on the plurality of images is reflected even when the imaging field of view is changed. On the other hand, a local influence that disappears when the imaging field of view is changed, such as a shadow of a foreign substance attached to the slide glass SG (for example, noise N4) is not reflected.
Thereafter, the process returns to the main routine.

  In step S12, the microscope system 1 acquires a specimen image. More specifically, when the specimen image acquisition instruction unit 142 outputs an instruction to acquire a specimen image, the visual field change control unit 144 responds to this by changing the imaging visual field VF to the visual field changing part 105 in the region where the specimen SP is present. (See FIG. 3C). Subsequently, the imaging control unit 143 causes the image acquisition unit 104 to image the imaging field of view VF. The specimen image obtained in this way is temporarily stored in the storage unit 130.

In subsequent step S13, the image calibration unit 153 calibrates the sample image using the corrected blank image. More specifically, the image calibration unit 153 obtains each pixel value S (x, y, z) of the sample image to be calibrated from the storage unit 130, and the pixel value G ^ of the corrected blank image and the glass image The pixel value S ^ (x, y, z) of the corrected sample image is calculated by the following equation (6) using the average value G _bias (ave) of the bias components of all the pixels. Here, S ^ indicates that the symbol "^ (hat)" is attached to S.

Figure 12 shows the calibrated specimen image M _CR1 of the specimen image M _OR1 was calibrated by correcting the blank image M _CB1. Note that an image M _CB1 ′ illustrated in FIG. 11 corresponds to an image obtained by dividing the average value G _bias (ave) of the bias component by the pixel value G ^ of the corrected blank image. As shown in FIG. 11, in the corrected sample image M _CR1 , shading caused by the inclination of the slide glass SG observed in the sample image M _OR1 and noise derived from the observation optical system (noises N1, N2, N3, N5) ) Is calibrated.

  As described above, according to the first embodiment, it is possible to create a corrected blank image that reflects only shading caused by the inclination of the slide glass and noise derived from the observation optical system. Therefore, by using such a corrected blank image, only the noise derived from the shading caused by the inclination of the slide glass and the observation optical system is left with respect to the specimen image while leaving the local noise attached to the slide glass. Can be calibrated.

  In the first embodiment, since a white image is used when creating a corrected blank image, it is possible to suppress the influence of local noise such as the shadow of a foreign substance adhering to the slide glass. Therefore, it is possible to create a corrected blank image that can remove only noise components derived from the observation optical system among various noises without performing complicated calculation processing.

  In the first embodiment, when the glass image is captured, the depth of focus is set in accordance with the imaging of the sample image, so that shading caused by the inclination of the slide glass can be accurately calibrated.

(Modification 1-1)
Next, Modification 1-1 of Embodiment 1 will be described.
In step S113, the pixel value W ^ of the corrected blank image may be created using the following equation (7) instead of the above equation (5). In Expression (7), the bias component G _bias (x, y, z) of the glass image, the average value G _bias (ave) of the bias components of all the pixels in the glass image, and the bias of all the pixels in the white image The average value W _bias (ave) of the component and the noise component W _noise (x, y, z) of the white image are used. W ^ indicates that the symbol "^ (hat)" is attached on W.

Here, in Equation (6), the value obtained by dividing the average value W _bias (ave) of the bias component of the white image by the average value G _bias (ave) of the bias component of the glass image is multiplied by the bias component of the glass image. This is because the dynamic range of the white image and the glass image is made uniform.

In this case, the pixel value of the calibrated specimen image is given by the following equation (8).

(Modification 1-2)
Next, Modification 1-2 of Embodiment 1 will be described.
In steps S111 and S112, when separating a blank image into a bias component and a noise component, a rolling ball algorithm (Stanley Sternberg, “Biomedical Image Processing”) , IEEE Computer, January 1983).

Specifically, in order to apply the rolling ball algorithm to a microscope image, first, a hemispherical kernel is set. The radius of the kernel is preferably set so that noise is removed. Hereinafter, the pixel value having the maximum value in the kernel is defined as W _MAX . This kernel may be set in advance.
Thereafter, the converted pixel value W ′ is calculated by converting the pixel value W of each pixel of the microscope image according to the following equation.
W ′ = W _MAX −W

Subsequently, in the space where the converted pixel value W ′ is a component in the height direction and the component in the horizontal axis direction is the distance from the central pixel, the center of the kernel (the center of the hemisphere) is applied to the target pixel. The output value W _OUT defined by the following equation is calculated for all the pixels included in the range (that is, the peripheral pixels of the target pixel).
W _OUT = W '+ W _KER
W _{KER on} the right side of this equation is a kernel value.

The smallest value among the output values W _OUT calculated thereby is set as a bias component in the target pixel. The noise component is calculated by subtracting the bias component from the converted pixel value of the target pixel.

(Embodiment 2)
Next, a second embodiment of the present invention will be described.
FIG. 13 is a block diagram illustrating a microscope system according to the second embodiment. As shown in FIG. 13, the microscope system 2 according to the second embodiment includes an image processing device 21 instead of the image processing device 11 shown in FIG. The configuration of the microscope apparatus 10 is the same as that in the first embodiment.

The image processing apparatus 21 includes a storage unit 210, a control unit 220, and a calculation unit 230 instead of the storage unit 130, the control unit 140, and the calculation unit 150 shown in FIG.
In addition to various control programs and various setting information, the storage unit 210 stores an image processing program 211 for calibrating a specimen image using a corrected blank image created using a plurality of glass images.

  The control unit 220 includes a blank image acquisition instruction unit 221 including a glass image acquisition instruction unit 221a, a specimen image acquisition instruction unit 142, an imaging control unit 143, and a visual field change control unit 144. Among these, the operations of the specimen image acquisition instruction unit 142, the imaging control unit 143, and the visual field change control unit 144 are the same as those in the first embodiment.

  The calculation unit 230 includes an image separation unit 231 including a glass image separation unit 231a, a correction blank image creation unit 232 including a common noise component extraction unit 232a, and an image calibration unit 153. Among these, the operation of the image calibration unit 153 is the same as that of the first embodiment.

  Next, the operation of the microscope system 2 will be described. The overall operation of the microscope system 2 is the same as that shown in FIG. 4, and details of the processing for acquiring a plurality of corrected blank images (step S10) and the processing for generating corrected blank images (step S11) are the same as those in the first embodiment. Is different.

FIG. 14 is a flowchart showing details of the process for acquiring a plurality of corrected blank images.
First, the microscope system 2 executes the process of loop A including steps S201 to S202 N times. Here, N is 3 or more.

  Specifically, when the glass image acquisition instruction unit 221a outputs an instruction to acquire a glass image in step S201, the visual field change control unit 144 responds to this by changing the imaging visual field VF to the specimen SP. Is moved to a region on the slide glass SG where no exists (see FIG. 3B). In subsequent step S202, the imaging control unit 143 causes the image acquisition unit 104 to image the imaging visual field VF (that is, a glass image).

After executing the process of loop A N times, the process returns to the main routine. Thereby, N glass images are acquired.
Images M _{G1 to} M _G3 shown in FIG. 15 show an example of a plurality (N = 3) of glass images acquired in this way. These images M _{G1 to} M _G3 are temporarily stored in the storage unit 210.

FIG. 16 is a flowchart showing details of the correction blank image creation process.
First, the microscope system 2 executes the process of loop B including step S211 N times, that is, the number of glass images acquired in step S10.

In step S211, the glass image separating unit 231a reads out the pixel values W (x, y) of the respective pixels constituting the glass image from the storage unit 210, the bias component G _bias (x, y, z), and the noise component G. A process of separating _noise (x, y, z) is executed. The details of the processing in step S211 are the same as those in the first embodiment (see step S112 in FIG. 7). Alternatively, as described in Modification 1-2, the glass image may be separated into a bias component and a noise component by a rolling ball algorithm.

Thus, in the case of FIG. 15, for example, an image M _G1 bias component G _bias of _{~M G3 (x, y, z} ) and the image M _GB1 ~M _GB3 corresponding to the noise component of the image M _G1 ~M _G3 G _noise ( Images M _{GN1 to} M _GN3 corresponding to x, y, z) are acquired.

In step S212, the common noise component extraction unit 232a extracts a common noise component from the noise components of the plurality of glass images.
More specifically, the common noise component extraction unit 232a firstly selects noise components G _{noise (i)} (x, y, z) and G _{noise (j)} (x ₎ of the same pixel in two arbitrarily selected glass images. , Y, z) (i ≠ j) are compared. When the absolute value | G _{noise (i)} (x, y, z) −G _{noise (j)} (x, y, z) | is smaller than a predetermined threshold, these noises The components G _{noise (i)} (x, y, z) and G _{noise (j)} (x, y, z) are determined to be common, and the common noise component G ^ _noise (x, y, z) (= G _{noise (i)} (x, y, z)) is extracted. As the threshold value at this time, an average value of noise generated in the image acquisition unit 104 is used.

The common noise component extraction unit 232a performs such determination for all combinations of the glass images acquired in step S10.
For example, in the case of FIG. 15, in the combination of the image M _GN1 and the image M _GN2 , noises N11 and N14 and noises N13 and N16 are extracted as common noise components. _Further , in the combination of the image M _GN2 and the image M _GN3 , noises N14 and N18 and noises N15 and N19 are extracted as common noise components, respectively. Further, in the combination of the image M _GN3 and the image M _GN1 , noises N18 and N11 and noises N19 and N13 are extracted as common noise components, respectively.

By such processing, a noise component G ^ _noise (x, y, z) derived from the observation optical system can be extracted. Note that the noise N12, the noise N16, and the noise N17 that are not extracted by this determination can be determined to be local adhered to the slide glass SG.

In step S213, the correction blank image creation unit 232 uses the common noise component G ^ _noise (x, y, z) extracted in step S212 and the bias component G _bias (x, y, z) of the glass image. Thus, the pixel value G ^ (x, y, z) of the corrected blank image given by the following equation (9) is calculated. At this time, as the bias component G _bias (x, y, z) of the glass image, any one of the plurality of glass images may be used, or the average value of the bias components of all the glass images. May be used.
Thereafter, the process returns to the main routine.

As described above, according to the second embodiment, the noise component derived from the observation optical system can be calculated using only the glass image. Therefore, it is possible to extract a noise component that accurately reflects the amount of noise on the glass image.
Further, according to the second embodiment, since it is not necessary to acquire a white image, it is possible to reduce the movement amount of the holding unit 101a.

(Modification 2)
Next, a second modification of the second embodiment will be described.
In step S212, a common noise component may be calculated using the following equation (10) instead of the above-described equation (9). In Expression (10), N is the number of glass images.

(Embodiment 3)
Next, a third embodiment of the present invention will be described.
FIG. 17 is a block diagram showing a microscope system according to the third embodiment. As shown in FIG. 17, the microscope system 3 according to the third embodiment includes an image processing device 31 instead of the image processing device 11 shown in FIG. The configuration of the microscope apparatus 10 is the same as that in the first embodiment.

The image processing apparatus 31 includes a storage unit 310 and a calculation unit 320 instead of the storage unit 130 and the calculation unit 150 shown in FIG.
In addition to various control programs and various setting information, the storage unit 310 stores an image processing program 311 for calibrating a specimen image with a corrected blank image created using a bias model that approximates a bias component of a glass image.

  The calculation unit 320 includes an image separation unit 151, a corrected blank image creation unit 321, and an image calibration unit 153. Among these, the operations of the image separation unit 151 and the image calibration unit 153 are the same as those in the first embodiment.

  The correction blank image creation unit 321 includes a bias model calculation unit 321a that calculates a bias model that approximates the bias component of the glass image, and a correction blank image approximation unit 321b that performs an approximation operation using the bias model.

  Next, the operation of the microscope system 3 will be described. The overall operation of the microscope system 3 is the same as that shown in FIG. 4, and the details of the correction blank image creation process (step S11) are different from those of the first embodiment.

FIG. 18 is a flowchart showing details of the correction blank image creation process.
First, the microscope system 3 executes the process of loop C including step S301 N times.
In step S <b _> 301, the image separation unit 151 reads a blank image from the storage unit 310 and executes a process of separating the image into a bias component B _bias (x, y, z) and a noise component B _noise (x, y, z). . The details of the process in step S301 are the same as those in the first embodiment (see step S112 in FIG. 7). Alternatively, as described in Modification 1-2, the glass image may be separated into a bias component and a noise component by a rolling ball algorithm.

In step S302, the bias model calculation unit 321a calculates an approximate expression that approximates the bias component B _bias (x, y, z) of the glass image. More specifically, first, the bias model calculation unit 321a performs bilinear interpolation on the bias component B _bias (x, y, z).

Here, bilinear interpolation is, as shown in FIG. 19, four coordinates (x ₁ , y ₁ ), (x ₂ , y ₁ ), (x ₁ , y ₂ ) surrounding the interpolation target coordinates (x, y). , (X ₂ , y ₂ ) values Src (x ₁ , y ₁ ), Src (x ₂ , y ₁ ), Src (x ₁ , y ₂ ), Src (x ₂ , y ₂ ), 11) is a method for calculating the interpolation value Dst.

Therefore, the bias model calculation unit 321a applies this bilinear interpolation operation to the bias component B _bias (x, y, z) of the glass image. Here, the luminance value at the interpolation target coordinate (x, y) is I ^ (x, y, z), and the luminance value at the four surrounding coordinates of the interpolation target coordinate is I ₁₁ (x ₁ , y ₁ , z), I _{21. (x 2, y 1, z} ), I 12 (x 1, y 2, z), as _{I 22 (x 2, y 2} , z), representing the bilinear interpolation operation by the following equation (12). The symbol I ^ indicates that the symbol "^ (hat)" is attached to the symbol I. The operation symbol BI represents a bilinear interpolation operation.
Expression (12 ′) is a further simplified expression (12).

Hereinafter, this expression (12 ′) is used as an approximate expression of the bias component B _bias (x, y, z) in the glass image. As an interpolation method, bicubic interpolation may be performed in addition to bilinear interpolation. Moreover, you may approximate by not only a plane but a curved surface.

In subsequent step S303, the corrected blank image approximating unit 321b calculates an approximate expression B _bias ^ = (B _bias , x, y, z) of the bias component of the glass image and a noise component B _noise (x, y, z), the pixel value B ^ of the corrected blank image is calculated by the following equation (13). Note that the noise component B _noise (x, y, z) derived from the observation optical system may be acquired from the noise component of the white image as described in the first embodiment, or described in the second embodiment. As described above, it may be acquired from noise components of three or more glass images.
Thereafter, the process returns to the main routine.

  As described above, according to the third embodiment, the approximate amount of the bias component is used when creating the interpolation blank image, so that the amount of calculation can be reduced. Therefore, it is possible to reduce the data capacity of the bias component and the amount of memory used.

(Embodiment 4)
Next, a fourth embodiment of the present invention will be described.
FIG. 20 is a block diagram illustrating a microscope system according to the fourth embodiment. As shown in FIG. 20, the microscope system 4 according to the fourth embodiment includes a microscope device 40 and an image processing device 41 instead of the microscope device 10 and the image processing device 11 shown in FIG.

  The microscope apparatus 40 further includes a magnification changing unit 109 in addition to the configuration of the microscope apparatus 10 illustrated in FIGS. 1 and 2. The magnification changing unit 109 performs an operation of changing the magnification of the objective lens 102 by, for example, rotating the revolver 108 illustrated in FIG.

The image processing apparatus 41 includes a storage unit 410, a control unit 420, and a calculation unit 430 instead of the storage unit 130, the control unit 140, and the calculation unit 150 shown in FIG.
In addition to various control programs and various setting information, the storage unit 410 stores an image processing program 411 that calibrates a sample image obtained by imaging the sample with the magnification changed by the microscope apparatus 10 using a corrected blank image. ing.

In addition to the configuration of the control unit 140 shown in FIG. 1, the control unit 420 further includes a magnification change control unit 421 that controls the operation of the magnification change unit 109.
The computing unit 430 further includes a corrected blank image size changing unit 431 in addition to the configuration of the computing unit 150 shown in FIG. The correction blank image size changing unit 431 changes the size (size) of the correction blank image created by the correction blank image creation unit 152 according to the magnification of the specimen image.

  Next, the operation of the microscope system 4 will be described. FIG. 21 is a flowchart showing the operation of the microscope system 4. Note that steps S10 and S11 shown in FIG. 21 correspond to steps S10 and S11 shown in FIG.

  In step S41 subsequent to step S11, when the specimen image acquisition instruction unit 142 outputs an instruction to acquire a specimen image, the visual field change control unit 144 responds to the visual field change unit 105 with the imaging visual field VF as an imaging target. The specimen SP is moved to the existence area.

In step S <b> 42, the magnification change control unit 421 causes the magnification change unit 109 to change the magnification in response to an instruction from the sample image acquisition instruction unit 142.
In step S43, the imaging control unit 143 causes the image acquisition unit 104 to perform imaging of the specimen image. Thereby, as shown in FIG. 22, for example, a sample image M _OR2 enlarged with respect to the sample image M _OR1 is acquired. This sample image M _OR2 is temporarily stored in the storage unit 410.

In step S44, the corrected blank image size changing unit 431 changes the size of the corrected blank image created in step S11 by the bilinear interpolation calculation represented by the above equation (11) according to the magnification changed in step S42. . The pixel value B ^ (x ', y', z) of the new corrected blank image after the size change is the pixel value corresponding to the pixel value B ^ (x, y, z) of the corrected blank image before the size change. Using B ^ (x ′, y ′, z), it is given by equation (14). Here, the coordinates (x ′, y ′) represent coordinates in the blank image after the size change. The relationship between the coordinates (x, y) in the blank image before the size change and the coordinates (x ′, y ′) in the blank image after the size change is as follows.
y = (Scale / Scale ') x y''
Scale represents the magnification before the size change, and Scale ′ represents the magnification after the size change. In the equation (14), the symbol B ^^ indicates that two symbols "^ (hat)" are superimposed on the code B.
Note that bicubic interpolation may be used as an interpolation method at this time.
For example, an image M _CB3 shown in FIG. 22 is obtained by enlarging the corrected blank image M _CB1 having a size corresponding to the sample image M _OR2 by bilinear interpolation.

In subsequent step S45, the image calibration unit 153 uses the corrected blank image after the size change to calibrate the sample image captured by changing the magnification. The specimen image calibration method using the corrected blank image is the same as in the first embodiment. Thereby, it is possible to obtain a specimen image from which noise caused by the shading caused by the inclination of the slide glass and the observation optical system is removed.
For example, the image M _CB3 ′ shown in FIG. 22 corresponds to an image obtained by dividing the average value G _bias (ave) of the bias component by the pixel value G ^ of the corrected blank image. By integrating the image M _CB3 ′ and the enlarged image M _OR2 , a calibrated sample image M _CR2 is obtained.

  As described above, according to the fourth embodiment, even when the magnification of the specimen image is changed, shading caused by the inclination of the slide glass and noise derived from the observation optical system are removed from the enlarged specimen image. Can do. The fourth embodiment can be applied to a case where the magnification is changed by inner focus, rear focus, or electronic zoom, in addition to the replacement of the objective lens.

(Embodiment 5)
Next, a fifth embodiment of the present invention will be described.
FIG. 23 is a block diagram showing a microscope system according to the fifth embodiment. As shown in FIG. 23, the microscope system 5 according to the fifth embodiment includes an image processing device 51 instead of the image processing device 11 shown in FIG.

The image processing apparatus 51 includes a storage unit 510, a control unit 520, and a calculation unit 530 instead of the storage unit 130, the control unit 140, and the calculation unit 150 shown in FIG.
In addition to various control programs and various setting information, the storage unit 510 processes and joins a plurality of sample images obtained by partially multi-band imaging the sample SP by the microscope apparatus 10 (VS). An image processing program 511 for generating an image is stored.

  The control unit 520 includes a VS image acquisition instruction unit 521 instead of the specimen image acquisition instruction unit 142 illustrated in FIG. The VS image acquisition instructing unit 521 gives support to the imaging control unit 143 and the visual field change control unit 144 to partially image the specimen SP in order to generate a VS image. According to the instruction of the VS image acquisition instruction unit 521, for example, as shown in FIG. 24, the imaging target area A1 on the slide glass including the specimen SP is divided into a plurality of small areas P1, P2,... P1, P2,... Are sequentially imaged along the arrows shown in the figure. Thereby, the sample images corresponding to the small areas P1, P2,... Are sequentially stored in the storage unit 510.

  Further, the calculation unit 530 includes a VS image generation unit 531 in addition to the configuration of the calculation unit 150 illustrated in FIG. The VS image generation unit 531 generates a VS image by connecting the sample images corresponding to the small regions P1, P2,... After being captured by the microscope apparatus 10 and subjected to the calibration processing by the image calibration unit 153. . By the operation of the VS image generation unit 531, a sample image with good image quality corresponding to the entire imaging target area A1 shown in FIG. 24 is obtained.

  When the calculation unit 530 creates a correction blank image used for calibration of the sample image corresponding to each of the small regions P1, P2,..., The small region not including the sample SP (for example, the small regions P1, P2, etc.) ) May be used (that is, a glass image). Alternatively, the white area may be captured by temporarily removing the imaging area from the slide glass SG.

  As described above, according to the fifth embodiment, it is possible to acquire a VS image with good image quality from which shading caused by the inclination of the slide glass and noise derived from the observation optical system are removed.

  The present invention is not limited to the above-described embodiments and modifications as they are, and various inventions can be formed by appropriately combining a plurality of constituent elements disclosed in the embodiments. . For example, some components may be excluded from all the components shown in the embodiment. Or you may form combining the component shown in different embodiment suitably.

(Appendix 1)
An image processing apparatus that calibrates a specimen image acquired by performing imaging by matching an imaging field of view with a region where a specimen exists in a microscope,
A plurality of blank images acquired by performing imaging by matching the imaging field of view with an area where the specimen does not exist, and acquired by imaging at least the imaging field of view according to an area where a slide glass exists The pixel values of the pixels constituting each of the plurality of blank images are separated into a bias component corresponding to the inclination of the pixel value and a noise component corresponding to noise, for a plurality of blank images including a glass image to be processed An image separation unit to perform,
A correction blank image creating unit that creates a correction blank image having, as pixel value components, a bias component included in the glass image and a component calculated from a noise component included in each of the plurality of blank images;
An image calibration unit that calibrates the specimen image using the corrected blank image;
An image processing apparatus comprising:

(Appendix 2)
The plurality of blank images include, in addition to the glass image, a white image acquired by performing imaging by matching the imaging field of view with an area where a slide glass does not exist,
The correction blank image creation unit creates the correction blank image from the bias component of the glass image and the component calculated from the noise component of the glass image and the white image. The image processing apparatus according to appendix 1.

(Appendix 3)
The plurality of blank images includes a plurality of glass images;
The correction blank image creation unit includes a noise component extraction unit that extracts a common noise component from the noise component of each of the plurality of glass images, and the common noise component and the plurality of glass images respectively The image processing apparatus according to appendix 1, wherein the corrected blank image is created from at least one of the bias components.

(Appendix 4)
The noise component extraction unit, when a difference in noise component value in a corresponding predetermined pixel between two glass images of the plurality of glass images is equal to or less than a predetermined threshold value, The image processing apparatus according to appendix 3, wherein the value of the noise component is the value of the common noise component.

(Appendix 5)
The image processing apparatus according to appendix 3, wherein the corrected blank image creation unit sets an average value of noise components in pixels corresponding to the plurality of glass images as the value of the common noise component.

(Appendix 6)
Any one of Supplementary notes 1 to 5, wherein the image separation unit separates pixel values of pixels constituting each of the plurality of blank images into the bias component and the noise component using a median filter process. The image processing apparatus according to claim 1.

(Appendix 7)
Any one of Supplementary notes 1 to 5, wherein the image separation unit separates pixel values of pixels constituting each of the plurality of blank images into the bias component and the noise component by using rolling ball processing. The image processing apparatus according to claim 1.

(Appendix 8)
The correction blank image creation unit calculates a pixel value of a pixel constituting the correction blank image from an approximate value approximating the bias component of the glass image and a noise component of each of the plurality of blank images. The image processing apparatus according to any one of appendices 1 to 7, wherein:

(Appendix 9)
A correction blank image size changing unit that changes the size of the correction blank image by performing an interpolation process based on the magnification of the microscope;
The said image calibration part calibrates the said sample image using the said correction blank image by which the magnitude | size was changed by the said correction blank image size change part, The appendix 1-8 characterized by the above-mentioned. Image processing device.

(Appendix 10)
At least one of the image separation unit, the correction blank image creation unit, and the image calibration unit executes processing for each band of the sample image, An image processing apparatus according to 1.

(Appendix 11)
The image processing apparatus according to any one of appendices 1 to 10,
A stage on which the slide glass can be placed;
An optical system provided facing the stage;
An image acquisition unit that acquires an image by imaging the imaging field through the optical system;
A visual field changing unit that changes the imaging visual field by moving at least one of the stage and the optical system in a direction orthogonal to the optical axis of the optical system;
A microscope having
A microscope system comprising:

(Appendix 12)
The microscope is caused to perform imaging by aligning the imaging field of view with an area where the specimen does not exist and acquire a blank image, and also perform imaging while acquiring the specimen image according to the area where the specimen is present. The microscope system according to appendix 11, further comprising a control unit that performs control.

(Appendix 13)
The microscope system according to appendix 12, wherein the control unit causes the microscope to capture the specimen image and the blank image at the same depth of focus.

(Appendix 14)
The control unit causes the microscope apparatus to image the specimen while relatively moving the stage on which the specimen is placed and the optical system in a direction perpendicular to the optical axis of the optical system, Take multiple specimen images,
The image calibration unit calibrates each of the plurality of sample images,
14. The microscope system according to appendix 12 or 13, wherein the image processing apparatus further includes a virtual slide generation unit that generates a virtual slide image obtained by connecting the calibrated specimen images to each other.

(Appendix 15)
The microscope system according to any one of appendices 11 to 14, wherein the microscope apparatus further includes a magnification changing unit that changes a magnification of an image to be captured.

(Appendix 16)
An image processing method for calibrating a sample image acquired by imaging a region including a sample in a microscope,
A plurality of blank images acquired by performing imaging by matching the imaging field of view with an area where the specimen does not exist, and acquired by imaging at least the imaging field of view according to an area where a slide glass exists The pixel values of the pixels constituting each of the plurality of blank images are separated into a bias component corresponding to the inclination of the pixel value and a noise component corresponding to noise, for a plurality of blank images including a glass image to be processed Image separation step to
A correction blank image creation step of creating a correction blank image having a bias component included in the glass image and a component calculated from a noise component included in each of the plurality of blank images as a pixel value component;
An image calibration step of calibrating the specimen image using the corrected blank image;
An image processing method comprising:

(Appendix 17)
An image processing program for causing a computer to execute a process of calibrating a specimen image acquired by imaging a region including a specimen in a microscope,
A plurality of blank images acquired by performing imaging by matching the imaging field of view with an area where the specimen does not exist, and acquired by imaging at least the imaging field of view according to an area where a slide glass exists The pixel values of the pixels constituting each of the plurality of blank images are separated into a bias component corresponding to the inclination of the pixel value and a noise component corresponding to noise, for a plurality of blank images including a glass image to be processed Image separation step to
A correction blank image creation step of creating a correction blank image having a bias component included in the glass image and a component calculated from a noise component included in each of the plurality of blank images as a pixel value component;
An image calibration step of calibrating the specimen image using the corrected blank image;
An image processing program comprising:

DESCRIPTION OF SYMBOLS 1-5 Microscope system 10, 40 Microscope apparatus 100 Arm 101 Stage 101a Holding part 102 Objective lens 103 Lens tube 104 Image acquisition part 105 Field of view change part 105a Motor 105b Scan part 106 Trinocular tube unit 107 Eyepiece unit 108 Revolver 109 Magnification change Unit 11, 21, 31, 41, 51 Image processing device 110 Input unit 120 Display unit 130, 210, 310, 410, 510 Storage unit 131, 211, 311, 411, 511 Image processing program 140, 220, 420, 520 Control 141, 221 Blank image acquisition instruction unit 141a White image acquisition instruction unit 141b, 221a Glass image acquisition instruction unit 142 Specimen image acquisition instruction unit 143 Imaging control unit 144 Field of view change control unit 150, 230, 320, 43 530 Calculation unit 151, 231 Image separation unit 151a White image separation unit 151b, 231a Glass image separation unit 152, 232, 321 Correction blank image creation unit 153 Image calibration unit 232a Common noise component extraction unit 321a Bias model calculation unit 321b Correction blank Image approximation unit 421 Magnification change control unit 431 Correction blank image size change unit 521 VS image acquisition instruction unit 531 VS image generation unit

Claims (17)

A storage unit that stores a plurality of sample images captured in an imaging field of view of a region where the sample exists and blank images captured in an imaging field of view of the region where the sample does not exist,
Of the blank images stored in the storage unit, for at least a plurality of blank images including a glass image captured in an imaging field of view where a slide glass exists, pixels of each of the plurality of blank images An image separation unit that separates a pixel value into a bias component corresponding to the slope of the pixel value and a noise component corresponding to noise;
A correction blank image creating unit that creates a correction blank image having, as pixel value components, a bias component included in the glass image and a component calculated from a noise component included in each of the plurality of blank images;
An image calibration unit that calibrates the specimen image using the corrected blank image;
An image processing apparatus comprising: The plurality of blank images include, in addition to the glass image, a white image imaged in an imaging field of view where the slide glass does not exist,
The correction blank image creation unit creates the correction blank image from the bias component of the glass image and the component calculated from the noise component of the glass image and the white image. The image processing apparatus according to claim 1. The plurality of blank images includes a plurality of glass images;
The correction blank image creation unit includes a noise component extraction unit that extracts a common noise component from the noise component of each of the plurality of glass images, and the common noise component and the plurality of glass images respectively The image processing apparatus according to claim 1, wherein the corrected blank image is created from at least one of the bias components.   The noise component extraction unit, when a difference in noise component value in a corresponding predetermined pixel between two glass images of the plurality of glass images is equal to or less than a predetermined threshold value, The image processing apparatus according to claim 3, wherein a value of the noise component is the value of the common noise component.   The image processing apparatus according to claim 3, wherein the correction blank image creation unit sets an average value of noise components in pixels corresponding to the plurality of glass images as the value of the common noise component.   The said image separation part isolate | separates the pixel value of the pixel which comprises each of these blank images into the said bias component and the said noise component using a median filter process. The image processing apparatus according to any one of the above.   The said image separation part isolate | separates the pixel value of the pixel which comprises each of these blank images into the said bias component and the said noise component using a rolling ball process. The image processing apparatus according to any one of the above.   The correction blank image creation unit calculates a pixel value of a pixel constituting the correction blank image from an approximate value approximating the bias component of the glass image and a noise component of each of the plurality of blank images. The image processing apparatus according to claim 1, wherein the image processing apparatus is an image processing apparatus. A correction blank image size changing unit that changes the size of the correction blank image by performing an interpolation process based on the magnification of the microscope;
The said image calibration part calibrates the said sample image using the said correction | amendment blank image by which the magnitude | size was changed by the said correction | amendment blank image size change part, The any one of Claims 1-8 characterized by the above-mentioned. Image processing apparatus.   10. The method according to claim 1, wherein at least one of the image separation unit, the correction blank image creation unit, and the image calibration unit executes processing for each band of the sample image. The image processing apparatus according to item. The image processing apparatus according to any one of claims 1 to 10,
A stage on which the slide glass can be placed;
An optical system provided facing the stage;
An image acquisition unit that acquires an image by imaging the imaging field through the optical system;
A visual field changing unit that changes the imaging visual field by moving at least one of the stage and the optical system in a direction orthogonal to the optical axis of the optical system;
A microscope having
A microscope system comprising:   The microscope is caused to perform imaging by aligning the imaging field of view with an area where the specimen does not exist and acquire a blank image, and also perform imaging while acquiring the specimen image according to the area where the specimen is present. The microscope system according to claim 11, further comprising a control unit that performs control.   The microscope system according to claim 12, wherein the control unit causes the microscope to image the sample image and the blank image with the same depth of focus. The control unit causes the microscope apparatus to image the specimen while relatively moving the stage on which the specimen is placed and the optical system in a direction perpendicular to the optical axis of the optical system, Take multiple specimen images,
The image calibration unit calibrates each of the plurality of sample images,
The microscope system according to claim 12 or 13, wherein the image processing apparatus further includes a virtual slide generation unit that generates a virtual slide image obtained by connecting the calibrated specimen images to each other.   The microscope system according to claim 11, wherein the microscope apparatus further includes a magnification changing unit that changes a magnification of an image to be captured. A storage step of storing a plurality of sample images captured in an imaging field of view of a region where the sample exists and blank images captured in an imaging field of view of the region where the sample does not exist;
Of the blank images stored in the storage step, at least for a plurality of blank images including a glass image captured in an imaging field of view where a slide glass exists, pixels of each of the plurality of blank images An image separation step of separating the pixel value into a bias component corresponding to the slope of the pixel value and a noise component corresponding to noise;
A correction blank image creation step of creating a correction blank image having a bias component included in the glass image and a component calculated from a noise component included in each of the plurality of blank images as a pixel value component;
An image calibration step of calibrating the specimen image using the corrected blank image;
An image processing method comprising: A storage step of storing a plurality of sample images captured in an imaging field of view of a region where the sample exists and blank images captured in an imaging field of view of the region where the sample does not exist;
Of the blank images stored in the storage step, at least for a plurality of blank images including a glass image captured in an imaging field of view where a slide glass exists, pixels of each of the plurality of blank images An image separation step of separating the pixel value into a bias component corresponding to the slope of the pixel value and a noise component corresponding to noise;
A correction blank image creation step of creating a correction blank image having a bias component included in the glass image and a component calculated from a noise component included in each of the plurality of blank images as a pixel value component;
An image calibration step of calibrating the specimen image using the corrected blank image;
An image processing program for causing a computer to execute.